refactor
Pass
Audited by Gen Agent Trust Hub on Aug 15, 2026
Risk Level: SAFECOMMAND_EXECUTION
Full Analysis
- [SAFE]: No malicious patterns, obfuscation, or data exfiltration vectors were detected. The skill follows software engineering best practices by requiring user confirmation and test verification for all changes.
- [COMMAND_EXECUTION]: The skill facilitates the execution of standard build and testing tools (such as npm, pytest, and mvn) and provided local analysis scripts (
analyze-complexity.py,detect-smells.py). These executions are scoped to the primary purpose of code analysis and refactoring within the developer's environment. - [PROMPT_INJECTION]: The skill is designed to process untrusted source code provided by the user. While this constitutes a surface for indirect prompt injection, the risk is addressed by the workflow's requirement for the agent to present findings and plans for manual user review and approval at each phase of the refactoring process.
Audit Metadata